• methotrexate
• pharmacogenetics
• rheumatoid arthritis

We read with interest the paper by Ranganathan et al proposing that pharmacogenetics may be a useful tool to help predict methotrexate (MTX) toxicity and efficacy in rheumatoid arthritis (RA).¹

One aspect they highlight is the potential role of drug efflux mechanisms in contributing to the lack of response to MTX in some patients. It is important to note that although they discuss the drug efflux transporter P-glycoprotein (P-gp) as being of interest, the paper they cite in support of this view actually reports an experiment in which MTX resistance was mediated by a different drug transporter, multidrug resistance protein 1 (MRP1).² A range of efflux transporters have been described, including P-gp, MRP, and breast cancer resistance protein (BCRP). Different drugs appear to be substrates for different efflux transporters.³ The drug transporter that mediates MTX resistance remains somewhat controversial.

Llorente et al studied 16 patients with RA and found higher P-glycoprotein (P-gp) levels in patients who were defined as being refractory to disease modifying drug treatment than in treatment responders.⁴ Similarly, Norris et al demonstrated increased P-gp expression in leukaemic cell lines resistant to methotrexate.⁵ In contrast, a study using mdr1 transgenic mice (which overexpress P-gp) showed they remain susceptible to methotrexate.⁶ Others have suggested that MTX may only become a substrate for P-gp when it enters cells by passive diffusion.⁷

To examine the effect of P-gp expression on MTX response we recently studied 20 patients with RA who were taking parenteral MTX at a stable dose for at least eight weeks. We compared P-gp expression on peripheral blood lymphocytes (PBLs) of patients with RA with those of 10 healthy controls.⁸ The patients had established RA, with a mean (SD) age and disease duration of 57.7 (9.7) and 15.9 (12.9) years respectively. Eighteen (90%) were seropositive, and 14 (70%) had been treated with ⩾3 drugs which had failed. The median (range) MTX dose and disease activity score (DAS)28 at study entry were 17.5 mg (range 7.5–25) weekly and 4.5 (range 1.8–6.7) respectively. PBLs were separated by gradient centrifugation. P-gp expression was measured using a monoclonal antibody directed to an external epitope of P-gp (UIC2). Samples were fixed and analysed by flow cytometry. The percentage positive P-gp cells were calculated using Cellquest software. No significant difference was seen in P-gp expression between patients with RA and healthy controls. Within the RA group, response to MTX (as measured by the DAS score) was not associated with P-gp expression and there was no significant difference between responders (DAS28 <3.7) and non-responders to MTX (DAS28 >3.7)⁸ (Kruskal-Wallis, p=0.27) (table 1).

These results support the view that MTX is not a P-gp substrate and that P-gp expression on PBLs is not associated with MTX response in RA. As far as we know, to date there have been no published studies examining other efflux transporters and clinical response to MTX in RA. Laboratory studies, however, suggest that other transporters particularly MRP1 and MRP3⁹ and BCRP¹⁰ are primarily involved in MTX efflux.

We therefore agree with Ranganathan et al that drug efflux transporters may contribute to the response to MTX in RA.¹ We also agree that genetic variability in the expression of such transporters may explain part of the heterogeneity of treatment response. The evidence to date, including our own observations, does not, however, support a significant role of P-gp in mediating this. Further studies are therefore required to determine which other drug transporters are most important in RA in determining MTX resistance. Given the central role played by MTX as a disease modifying antirheumatic drug in RA, modulation of any such transporters using specific chemosensitising agents may provide a new and rational additional intervention in patients with RA.